In a magnetic recording disk drive, data is written by thin film magnetic transducers called "heads", which are supported over a surface of the disk while it is rotated at a high speed. The heads are supported by a thin cushion of air (an "air bearing") produced by the disk's high rotational speed.
A prior art "merged" magnetoresistive (MR) read/inductive write head is shown in the side sectional view of FIG. 1 and the partial end view, as seen from the disk, of FIG. 2. The term "merged" read/write head means that a magnetoresistive read element (or head) is located between first and second magnetic shields and an inductive write element (or head) has first and second pole tips spaced by a nonmagnetic gap, where the second magnetic shield for the magnetoresistive read element also serves as the first pole tip for the inductive write element. The thin film inductive write head includes bottom and top pole pieces P1 and P2, respectively, that are formed from thin films ("layers") of magnetic material. The pole pieces have a pole tip height dimension commonly called "throat height". In a finished write head, throat height is measured between an air-bearing surface ("ABS"), formed by polishing the tips of the pole pieces, and a zero throat height level ("zero throat level"), where the bottom pole piece P1 and the top pole piece P2 converge at the magnetic recording gap G. A thin film magnetic write head also includes a "pole tip region" which is located between the ABS and the zero throat level, and a "back region" which extends back from the zero throat level to and including a back gap BG. Each pole piece has a pole tip portion in the pole tip region and a back portion in the back region. The pole pieces are connected together at the back gap BG. The pole tips are extensions of the bottom and top pole pieces P1 and P2 of the write head. Each of the pole pieces P1 and P2 transitions to a pole tip in the pole tip region. The pole tips are separated by a gap G, which is a thin layer of insulation material, typically alumina (Al.sub.2 O.sub.3). The pole tip of the top pole piece P2 is the last element to induce flux into the magnetic layer on the disk; therefore, its width is more important than the width of the pole tip on the bottom pole piece P1. However, it is important for the pole tips to have the same width to minimize stray flux leakage around the gap.
A merged MR head, such as shown in FIGS. 1 and 2, employs an MR read element and an inductive write element in combination. This is accomplished by using the top shield S2 of the MR element as the bottom pole P1 of the write element. A merged MR head has a high capability for either reading or writing. The merged MR head saves processing steps over constructing separate read and write heads because the second shield layer S2 of the MR read head also serves as the bottom pole P1 for the write head, thereby eliminating a fabrication step.
However, merged MR head structures generate significantly large side-fringing fields during recording. These fields are caused by flux leakage from the top pole P2 to the parts of the bottom pole P1 beyond the region defined by P2. The side-fringing fields limit the minimum trackwidth achievable, and therefore limit the upper reach of track density. Consequently, when a track written by the write element of a merged MR head is read by the MR element, the "off-tracK" performance of the MR element is poor. That is, when the MR element is moved laterally from the center of a track being read, it cannot move far before interference from the fields of the adjacent track begins to interfere with the fields of the track being read.
One solution to the side-fringing problem of the merged MR head is to construct a narrow "pedestal" pole tip portion PT1b on top of the second shield layer S2, as shown in FIG. 2, with the P1/S2 layer then serving as a wider bottom pole tip element PT1a. Both of these pole tip elements PT1b and PT1a form the pole tip portion of the bottom pole P1, with the pole tip element PT1b forming a pedestal on the pole tip element PT1a. The sidewalls of the bottom and top pole tips PT1b and PT2 are substantially vertically aligned and constrained to substantially equal widths.
The conventional process for fabricating the pole pieces and gap of the merged MR head involves sputter deposition and lithographic patterning. The pedestal PT1b pole tip element is formed by ion beam milling through the top and bottom pole pieces and gap layer G, using P2 as a mask, to remove the gap layer G and a portion of layer P1 in the region not beneath P2. It is this ion beam milling with P2 as a mask that causes the sidewalls of the bottom and top pole tips PT1b and PT2 to be substantially vertically aligned. Because the PT1b pole tip element is formed by ion beam milling of the gap layer G and a portion of the P1 layer not beneath the P2 layer, this type of P1 pole tip is also called a "notched" pole tip. However, because of shadowing of the ion beam caused by the top pole tip PT2 during the ion beam milling process, there is some outward taper to the bottom pole tip PT1b. In addition, because the ion beam milling rate of the material of gap layer G is slower than the ion beam milling rate of the Ni--Fe material of P1, and because P2 is to serve as the mask, a much thicker layer of P2 is required to compensate for the loss of P2 material during the ion beam milling.
What is needed is process for forming a thin film inductive write head, and particularly a merged magnetoresistive read/inductive write head, that does not require a substantially thicker P2 layer.